Spotlight on Eigenfactor Metrics

Everyone knows JCR because of the Journal Impact Factor, the journal evaluation metric based on citations. But the JCR is so much more than this one indicator. Another key indicator in the JCR is the Eigenfactor, developed by the DataLab at the University of Washington, which takes a network approach to journal evaluation. The State of Innovation takes a closer look at the Eigenfactor in this article.

The DataLab’s main goal is to answer the question: We have big data – what do we do with it? In terms of journals, librarians need to manage collections on a budget, publishers need to promote their portfolios of journals, foundations and societies need to know their impact, government agencies need to know how to allocate funding as well as measure and demonstrate their impact on society, and aggregators need to know how to best structure and organize their data.

The algorithm for the Eigenfactor is based on an iterative voting scheme, or a “random walk,” around the citation network, with the scores being based on the amount of time a user is likely to spend in a given journal in the network. The Eigenfactor is set up so that all the scores within the JCR add up to 100, which means that the majority of the journals in the JCR have fractional scores.

Because these fractional scores tend to be difficult for non-power users to ingest, the DataLab developed the Normalized Eigenfactor, which is essentially the Eigenfactor expressed as a whole number. This is a simple normalization in which the ordinal ranks are exactly the same, there are no shifts in the relative rankings of the journals, and the ratios between any two journals do not change.

eigenfactor-Comparison-chart
With the Normalized Eigenfactor, the average journal has a score of 1, so the journal Ecology in the chart above with a Normalized Eigenfactor score of 7.88 is 7.88 times as influential as the average journal in the whole of the JCR. Ecology’s Normalized Eigenfactor still places it at #3 in the above chart—the same as it would if the ranking were based on the Eigenfactor score.

The Normalized Eigenfactor also corrects for inflation; i.e. when the size of the corpus grows, the Normalized Eigenfactor scores adjust to reflect this change. The Normalized Eigenfactor also complements the Article Influence Score better than the Eigenfactor Score does, because the Article Influence Score has always done this sort of normalization at the article level.

Adding the suite of Eigenfactor metrics to Journal Impact Factor helps to build a more complete view of a journal’s mark in the academic realm.

Read more about the Eigenfactor.


Did you miss any of the posts in our JCR series? Catch up now:
The 2016 Journal Citation Reports Release Is Coming!
Behind the Scenes of the JCR Selection & Production Processes
Spotlight on: Downloadable Reports in JCR
Spotlight on: Visualizations in Journal Citation Reports
Journal Suppressions in the 2015 JCR Data - Why So Few?
Recap of the 2016 Journal Citation Reports Release
Beyond the Journal Impact Fact